Bounded Distrust
Read the original on Astral Codex Ten →
Summary
Develops the influential concept of 'bounded distrust': biased media and expert establishments operate within rules — there are lines they cross (spin, misleading-but-literally-true framing, selective context, suppressing inconvenient research) and lines they won't (FOX won't fake a mass shooting with deepfakes; experts won't sign a flatly-false petition). So even a source you rightly distrust can be trusted on specific things if you know the rules of the lying game. He walks a liberal trusting FOX on a shooting, a conservative trusting WaPo on the 2020 election (vs the misleading Lincoln-Marx human-interest piece), and the Swedish immigrant-crime researchers (the establishment punishes but won't directly assert the false claim) — concluding that because experts won't sign a transparently false statement, the existence of a climatologist consensus petition on AGW is itself meaningful evidence. Caps with the ivermectin coda and the savvy-vs-clueless dynamic (the Soviet 'good harvest,' Dan Quayle's 'revenue enhancement'): savvy insiders read the rules effortlessly, clueless people can't separate the cases where media lies from where it doesn't, and the two types need to understand each other.
Why this score
Quality 82 · Excellent. Excellent band, high. One of his important, widely-cited epistemics essays: it names and clarifies a genuinely useful framework (extract truth from biased sources by learning which lines they won't cross), unifying media trust, expert trust, and conspiracism, with sharp exportable pieces (the consensus-petition argument; savvy-vs-clueless). Co-tier with the Cult Of Smart review (82); held below the very top because it is short and develops a single idea.
Claude’s paradigm shift 65 · Notable shift. Notable shift, upper. 'Bounded distrust' was a genuinely fresh and influential framing at publication; the savvy-vs-clueless dynamic and the consensus-petition inference are original contributions rather than restatements.
Real-world impact 3 · Moderate. Names and clarifies a genuinely useful, widely-cited epistemics framework ('bounded distrust' — extract truth from biased sources by learning which lines they won't cross), with sharp exportable pieces. Conceptual influence within rationalist/media-criticism discourse, no material change — modest RWI.